Atypical integration of multisensory inputs has been suggested as a major component of autism, with some clinical and behavior-based empirical support for this view. Where and when in the neural processing stream these sensory integration deficits occur is as yet unknown, and gaining an understanding of this will be essential in defining the neuropathology of autism. In fact, there is precious little understanding of the basic development of healthy sensory integration mechanisms in typically developing children, although recent work in animal models is beginning to shed some light. Under this project, we will use established electrophysiological metrics of multisensory integration that we have developed in our laboratory in healthy adults, to test the hypothesis that multisensory integration is impaired in autism. The high-density electrical recordings of neural activity that we record will provide a precise measure of when in the information processing stream sensory integration differs from typically developing matched controls, as well as a good model of the underlying brain processes that are affected. We expect that there are profound developmental effects on how multisensory inputs are treated and we will therefore also characterize the normal developmental trajectory of multisensory integration in typically developing children, using a cross-sectional approach. Specific hypotheses about when and where multisensory processes will be affected in autism are made based on the thesis that there is impoverished connectivity between distant cortical regions in this population, and our predictions are predicated on a rudimentary three-stage model of multisensory integration that we have developed in light of the extant literature. The data acquired under this project will provide a strong empirical test of deficits in multisensory integration processes in autism. Understanding how multisensory integration develops and changes over childhood will significantly inform models of multisensory integration, and provide an initial benchmark against which predictions about possible disordered multisensory integration in a host of developmental disorders can be made.